Synthesis of hypermodified adenosine derivatives as selective adenosine A3 receptor ligands

10
Synthesis of hypermodified adenosine derivatives as selective adenosine A 3 receptor ligands Liesbet Cosyn, a Zhan-Guo Gao, b Philippe Van Rompaey, a Changrui Lu, b Kenneth A. Jacobson b, * and Serge Van Calenbergh a, * a Laboratory for Medicinal Chemistry (FFW), UGent, Harelbekestraat 72, B-9000, Belgium b Molecular Recognition Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes and Digestive and Kidney Diseases, National Institutes of Health (NIH), DHHS, Bldg. 8A, Rm. B1 A-19, NIH, NIDDK, LBC, Bethesda, MD 20892-0810, USA Received 4 July 2005; revised 21 September 2005; accepted 27 September 2005 Available online 2 November 2005 Abstract—We investigated the A 3 AR affinity and selectivity of a series of 2-substituted 3 0 -azido and 3 0 -amino adenosine derivatives as well as some 5 0 -uronamide derivatives thereof. All compounds showed high A 3 AR selectivity. While the 3 0 -azides appeared to be A 3 AR antagonists with moderate A 3 AR affinity, their 3 0 -amino congeners exhibit significantly improved A 3 AR affinity and behave as partial agonists. For both the 3 0 -azides and the 3 0 -amines, the 5 0 -methylcarbamoyl modification improved the overall affinity. Introduction of a 2-phenylethynyl substituent provided high affinity for the A 3 AR. Ó 2005 Elsevier Ltd. All rights reserved. 1. Introduction Adenosine receptors (AR) belong to the family of G protein-coupled receptors. They are subdivided into four subtypes designated A 1 ,A 2A ,A 2B and A 3 , accord- ing to the chronological discovery of the receptors. 1 The A 3 ARs are coupled to G i proteins and, therefore, inhibit adenylate cyclase leading to a decrease in intracellular levels of cAMP. 2 The selective activation of the A 3 AR is both cardioprotective and cerebroprotective in a vari- ety of ischaemic models. 3,4 Selective A 3 AR antagonists promise to be useful in the regulation of cell growth 5,6 and as anti-asthmatic, 7 cerebroprotective 4,8 and anti- inflammatory agents. 9 A 3 AR antagonists appear to low- er the intraocular pressure in mice and monkeys and are proposed as new potential therapeutics for the treatment of glaucoma. 10,11 Adenosine receptors are ubiquitously distributed throughout the body. As a consequence, ligands need to be highly selective in their action with respect to receptor subtype and tissue to be of therapeutic value. 12 Numerous structure–activity studies of adenosine deri- vatives as receptor agonists conclude that selectivity may be provided by specific substitutions of the adenine ring. 13,1 Substitution at the 8-position of the ring is not well tolerated by any AR subtype. 14,15 The nitrogen atoms at positions 3 and 7 are required for high affinity of adenosine at all subtypes. 1 2-Alkynyl derivatives of NECA possess high affinity at the A 3 receptor subtype. Moreover, the presence of 2-alkyne substituents en- hanced the A 3 AR selectivity. 16 DeNinno et al. discovered that introduction of an amino group at the 3 0 position improves the selectivity for the human A 3 AR, while enhancing the water solubility. The affinity drop caused by this 3 0 -substitution could be overcome by elaborating the N 6 -substituents. 17 The combination of a large N 6 -substitituent with a 2-alkynyl group has proven to be unsuccessful because of the ste- ric hindrance caused by the two large substituents, reflected by a decrease in A 3 AR affinity. 16 Therefore, the present study investigated the effect of a 2-alkynyl substituent in concert with a small N 6 -substituent on the affinity and selectivity of a series of 3 0 -azido and 3 0 -amino adenosine derivatives. In addition, we evaluat- ed the effect of the 5 0 -methylcarbamoyl modification on the overall affinity and efficacy of these compounds (Fig. 1). 0968-0896/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.bmc.2005.09.062 Keywords: Adenosine A 3 receptor; Selectivity; Affinity; Ribose and purine modifications. * Corresponding authors. Tel.: +32(0)9 264 81 24; fax: +32(0)9 264 81 46 (S.V.C.); tel.: +1 301 496 9024; fax: +1 301 480 8422 (K.A.J.); e-mail addresses: [email protected]; serge.vancalenbergh@ ugent.be Bioorganic & Medicinal Chemistry 14 (2006) 1403–1412

Transcript of Synthesis of hypermodified adenosine derivatives as selective adenosine A3 receptor ligands

Bioorganic & Medicinal Chemistry 14 (2006) 1403–1412

Synthesis of hypermodified adenosine derivativesas selective adenosine A3 receptor ligands

Liesbet Cosyn,a Zhan-Guo Gao,b Philippe Van Rompaey,a Changrui Lu,b

Kenneth A. Jacobsonb,* and Serge Van Calenbergha,*

aLaboratory for Medicinal Chemistry (FFW), UGent, Harelbekestraat 72, B-9000, BelgiumbMolecular Recognition Section, Laboratory of Bioorganic Chemistry, National Institute of Diabetes

and Digestive and Kidney Diseases, National Institutes of Health (NIH), DHHS, Bldg. 8A, Rm. B1 A-19, NIH,

NIDDK, LBC, Bethesda, MD 20892-0810, USA

Received 4 July 2005; revised 21 September 2005; accepted 27 September 2005

Available online 2 November 2005

Abstract—We investigated the A3AR affinity and selectivity of a series of 2-substituted 3 0-azido and 3 0-amino adenosine derivativesas well as some 5 0-uronamide derivatives thereof. All compounds showed high A3AR selectivity. While the 3 0-azides appeared to beA3AR antagonists with moderate A3AR affinity, their 3 0-amino congeners exhibit significantly improved A3AR affinity and behaveas partial agonists. For both the 3 0-azides and the 3 0-amines, the 5 0-methylcarbamoyl modification improved the overall affinity.Introduction of a 2-phenylethynyl substituent provided high affinity for the A3AR.� 2005 Elsevier Ltd. All rights reserved.

1. Introduction

Adenosine receptors (AR) belong to the family of Gprotein-coupled receptors. They are subdivided intofour subtypes designated A1, A2A, A2B and A3, accord-ing to the chronological discovery of the receptors.1 TheA3ARs are coupled to Gi proteins and, therefore, inhibitadenylate cyclase leading to a decrease in intracellularlevels of cAMP.2 The selective activation of the A3ARis both cardioprotective and cerebroprotective in a vari-ety of ischaemic models.3,4 Selective A3AR antagonistspromise to be useful in the regulation of cell growth5,6

and as anti-asthmatic,7 cerebroprotective4,8 and anti-inflammatory agents.9 A3AR antagonists appear to low-er the intraocular pressure in mice and monkeys and areproposed as new potential therapeutics for the treatmentof glaucoma.10,11

Adenosine receptors are ubiquitously distributedthroughout the body. As a consequence, ligands needto be highly selective in their action with respect to

0968-0896/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.bmc.2005.09.062

Keywords: Adenosine A3 receptor; Selectivity; Affinity; Ribose and

purine modifications.* Corresponding authors. Tel.: +32(0)9 264 81 24; fax: +32(0)9 264 81

46 (S.V.C.); tel.: +1 301 496 9024; fax: +1 301 480 8422 (K.A.J.);

e-mail addresses: [email protected]; serge.vancalenbergh@

ugent.be

receptor subtype and tissue to be of therapeutic value.12

Numerous structure–activity studies of adenosine deri-vatives as receptor agonists conclude that selectivitymay be provided by specific substitutions of the adeninering.13,1 Substitution at the 8-position of the ring is notwell tolerated by any AR subtype.14,15 The nitrogenatoms at positions 3 and 7 are required for high affinityof adenosine at all subtypes.1 2-Alkynyl derivatives ofNECA possess high affinity at the A3 receptor subtype.Moreover, the presence of 2-alkyne substituents en-hanced the A3AR selectivity.16

DeNinno et al. discovered that introduction of an aminogroup at the 3 0 position improves the selectivity for thehuman A3AR, while enhancing the water solubility.The affinity drop caused by this 3 0-substitution couldbe overcome by elaborating the N6-substituents.17 Thecombination of a large N6-substitituent with a 2-alkynylgroup has proven to be unsuccessful because of the ste-ric hindrance caused by the two large substituents,reflected by a decrease in A3AR affinity.16 Therefore,the present study investigated the effect of a 2-alkynylsubstituent in concert with a small N6-substituent onthe affinity and selectivity of a series of 3 0-azido and3 0-amino adenosine derivatives. In addition, we evaluat-ed the effect of the 5 0-methylcarbamoyl modification onthe overall affinity and efficacy of these compounds(Fig. 1).

Figure 1.

1404 L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412

2. Results and discussion

2.1. Chemistry

3-Azido-3-deoxy-1,2-di-O-acetyl-a-DD-ribofuronamide (10)was prepared from the commercially available 1,2-O-isopropylidene-a-DD-xylofuranose as described by us pre-viously.18 It was coupled under Vorbruggen conditionswith silylated 2-amino-6-chloropurine to give 11 in79% yield. Classical procedures allowed a straightfor-ward conversion of 11 to 13. Triphenylphosphine reduc-tion of the azido moiety yielded the correspondingamine 7.

Based on the results of Cristalli et al.19 we have chosenphenylethynyl as the most promising C2-substituent.Reaction conditions used to perform a Sonogashiracoupling20 of 13 with phenylacetylene yielded the 3 0-(4-phenyl-1,2,3-triazol-1-yl) derivative (15) of the 2-alkyny-lated compound (Scheme 1). This result was due to a

Scheme 1. Reagents and conditions: (a) i—HMDS, (NH4)2SO4, reflux, 20

CH2I2 in THF, reflux; (c) CH3NH3Cl, Et3N, EtOH, reflux; (d) Ph3P, H2O,

Cu+-catalysed Huisgen dipolar cycloaddition21 of the3 0-azide with phenylacetylene. Consequently, anotherstrategy was used to gain access to compound 14(Scheme 2), starting from 6-chloro-2-iodo-(9-tetrahy-dropyran-2-yl)purine (16), obtained from the 2-unsub-stituted analogue via a lithiation-mediated stannyltransfer process followed by 2-tributylstannyl-iodine ex-change.22 Sonogashira coupling of 17, followed bydeprotection, provided 19. Unfortunately, classicalVorbruggen coupling,23 as described for the synthesisof 11, did not give satisfying results. By using N,O-bis(trimethylsilyl)acetamide (BSA) as silylating agent,24

9-(2-acetyl-3-azido-3-deoxy-5-methylcarbamoyl-b-DD-ribo-furanosyl)-N6-methyl-2-phenylethynyladenine (20) wasobtained in poor yield.

During the course of this work it became clear that the3 0-amino-analogues generally exhibit much better A3ARaffinities than their 3 0-azide precursors. Consequently,we focussed only on the 3 0-amino derivatives for further

h, ii—2-amino-6-chloropurine, TMSOTf; (b) isoamyl nitrite, I2, CuI,

THF, 2 days; (e) phenylacetylene, (PPh3)2PdCl2, CuI, Et3N, DMF.

Scheme 3. Reagents and conditions: (a) 2-amino-6-chloropurine, BSA, TMSOTf, CH3CN, reflux; (b) isoamylnitrite, I2, CuI, CH2I2 in THF, reflux;

(c) i—CH3NH3Cl, Et3N, EtOH, reflux, ii—7 N NH3 in MeOH; (d) Na� in MeOH; (e) Ph3P, H2O, THF, 2 days; (f) alkyne, (PPh3)2PdCl2, CuI, Et3N,

DMF.

Scheme 2. Reagents and conditions: (a) CH3NH3Cl, DMAP, EtOH, reflux; (b) phenylacetylene (PPh3)2PdCl2, CuI, Et3N, DMF; (c) TFA, CH2Cl2;

(d) 10, BSA, TMSOTf, CH3CN reflux; (e) 7 N NH3 in MeOH; (f) Ph3P, H2O, THF, 2 days.

L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412 1405

synthesis. This direction permitted us to reduce the 3 0-azide to a 3 0-amine before the Sonogashira coupling,avoiding the unwanted cycloaddition.

9-(3-Amino-3-deoxy-b-DD-ribofuranosyl)-N6-methyl-2-iodo-purine (1) served as a suitable synthon for the synthesisof the 2-alkynylated 3 0-amino-adenosines 3–5 (Scheme3). It was obtained by coupling of sugar 21 with 2-ami-no-6-chloropurine. Elaboration of the base moiety toyield 25 was essentially accomplished as for 13. Stau-dinger reduction allowed the unmasking of the aminegroup. Finally, Sonogashira coupling on amine 1 pro-vided the alkynylated analogues 3–5 in 80–82% yield.

To continue the exploration of the 2-position, wesynthesized the 2-I and the 2-NH2 derivatives of the5 0-OH and the 2-H and the 2-I derivatives of the5 0-methylcarbamoyl 3 0-amino-N6-aminomethyl adeno-sine analogues (Scheme 4).

3 0-Amine 6 (Fig. 1) was prepared by catalytic hydroge-nation of the 3 0-azide precursor which has beendescribed.18

2.2. Biological evaluation

For the adenosine derivatives prepared in this study(1–9, 13, 14, 15, 25, 27 and 29) we measured boththe binding affinities at the hA1, hA2A and hA3ARand their degrees of activation of the A3AR subtype.The results are reported in Table 1. The ability of eachof these adenosine derivatives to compete for radioli-gand binding at each of these hARs was evaluated ata fixed concentration of 10 lM, and full competitioncurves were determined at the A3AR. Six different 2-substituents were included: H, I, NH2, Ph-C„C,pMePh-C„C and nBu-C„C. The choice of the methylgroup as a small N6 substituent was based on theresults of Cristalli et al., who demonstrated that it

Table 1. Binding affinities of adenine derivatives at human A1, A2A and A3ARs expressed in CHO cellsa

N

NN

N

N

O

OHR3

R2

R1

H Me

Compound R1 R2 R3 % Inhibition (hA3AR) Ki (nM) % Activationd (hA3AR)

(hA1AR) (%) (hA2AAR) (%)

1 I CH2OH NH2 39 12 879 ± 346 67 ± 6

2 NH2 CH2OH NH2 16 4 654 ± 42 57 ± 2

3 Ph-C„C CH2OH NH2 50 14 126 ± 4 36 ± 6

4 pMePh-C„C CH2OH NH2 33 0 145 ± 35 23 ± 3

5 nBu-C„C CH2OH NH2 4 3 389 ± 112 66 ± 11

6 H CONHMe NH2 18 12 32.3 ± 3.9 72 ± 10

7 I CONHMe NH2 16 24 71.4 ± 12.8 18 ± 10

8 NH2 CONHMe NH2 �14 10 536 ± 247 92 ± 3

9 Ph-C„C CONHMe NH2 9 9 15.6 ± 3.6 67 ± 11

13 I CONHMe N3 16 0 2530b �8 ± 5

14 Ph-C„C CONHMe N3 31 16 78.9 ± 12.4 �11 ± 8

15 Ph-C„C CONHMe 4-Ph-1,2,3-triazol-1-yl 14 2 1820 ± 770 0

25 I CH2OH N3 85b,c 27% 6540 ± 320 �3 ± 2

27 NH2 CH2OH N3 49 16 28,800b 0

2918 H CONHMe N3 12 10 1140 ± 300 38 ± 4

a All A3AR experiments were performed using adherent CHO cells stably transfected with cDNA encoding one of the human adenosine receptors.

Binding at human A1, A2A and A3ARs in this study was carried out as described in methods using [3H]PIA, [3H]CGS 21680 or [125I]AB-MECA as

radioligand. Values from the present study are expressed as Ki values (means ± SEM, n = 3, unless noted) or as percent displacement of radioligand

at 10 lM.b n = 1.cKi (A1 AR) = 1850 nM.d% Inhibition at 100 lM of forskolin-stimulated cAMP production at 10 lM, in CHO cells expressing the hA3AR, as a percentage of the response of

the full angonist CI-IB-MECA (n = 3).

Scheme 4. Reagents and conditions: (a) i—CH3NH3Cl, Et3N, EtOH, reflux, ii—7 N NH3 in MeOH; (b) Na� in MeOH; (c) Ph3P, H2O, THF, 2 days.

1406 L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412

increased the affinity for the human A3AR and signif-icantly enhanced the A3AR selectivity.19

Results from the competition experiments showed thatthe A3AR affinities of the 3 0-amines were much higher

than those of the 3 0-azides. For all derivatives studied,the 5 0-methylcarbamoyl modification, in general, en-hanced the affinity at the A3AR in comparison to 5 0-CH2OH. Except for 25, all evaluated compounds showeda very high selectivity for theA3ARcompared to the other

L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412 1407

ARs. The most potent compound (9) displayed aKi valueof 16 nM at the A3AR. The C-2 substituent of this com-pound, a phenylethynyl moiety, was previously shownto enhance A3AR affinity and selectivity,19 and provedto have a superior contribution to A3AR affinity than ap-methyl-phenylethynyl (4) or a 1-hexynyl (5) moiety.Furthermore, the 2-phenylethynylmodification appearedto overcome the reduction of affinity caused by the3 0-azide (cf.Ki = 78.9 nM for 14 vs 1140 nM for 29). Con-sequently, this C-2 substituent was selected to be com-bined with a 3 0-amino and a 5 0-methylcarbamoylmodification.

Previous studies showed that 3 0-amino derivatives exhibita decreased affinity at the A3AR compared to their 3 0-hy-droxy analogues. The affinity reduction associated withthis 3 0 modification could be overcome by elaboratingtheN6-substituents, for examplewith a substituted benzylgroup.17 The high affinity of compound 9 (Ki = 15.6 nM)demonstrated that a 2-phenylethynylmodification in con-cert with a small N6-substituent was likewise capable ofovercoming this reduction in affinity. Note that in ourexperiments the affinity of derivative 6 for the A3AR(Ki = 32.3 nM) was 4-fold higher than that reported byDeNinno et al.15 The 2-I analogue 7 also showed appre-ciable A3AR affinity (Ki = 71.4 nM). Conversely, the 2-NH2 analogue 8 exhibited weak A3AR affinity(Ki = 536 nM).

The results of the cyclic cAMP-assay (Table 1) indicatedthat all 3 0-azides were A3AR antagonists, except forcompound 29 which showed partial agonist activity.Also the 3 0-(4-phenyl-1,2,3-triazol-1-yl) derivative 15 ap-peared to be an A3AR antagonist. All other compoundswere partial agonists, except for compound 8, whichmanifested full agonist activity.

3. Conclusions

The 2,3 0,5 0-trisubstituted and 2,3 0-disubstituted N6-methyl adenosine derivatives described in the presentstudy were synthesized in good overall yields. All thecompounds had A3AR affinities in the low micromolaror nanomolar range and showed very high A3AR selec-tivity. The 3 0-azides appeared to be A3AR antagonistswith a moderate A3AR affinity. The 3 0-amino modifica-tion significantly improved the A3AR affinity and result-ed in partial A3AR agonists. For both the 3 0-azido andthe 3 0-amino derivatives, the 5 0-methylcarbamoyl modi-fication improved the overall affinity. Curiously, thepresence of a 5 0-uronamide did not restore full A3ARefficacy in 2-position derivatives, as was demonstratedin the case of N6-substituents that reduced efficacy.26

The 2-phenylethynyl derivative 9 demonstrated highA3AR receptor affinity with a Ki value of 15.6 nM and>1000-fold selectivity. Previous studies revealed that3 0-amines exhibit a decreased affinity compared to their3 0-hydroxy analogues. This study demonstrated thatintroduction of a 2-phenylethynyl substituent in concertwith the N6-methyl group is capable of overcoming thisaffinity drop.

4. Experimental

4.1. Chemicals and solvents

All reagents were from standard commercial sourcesand of analytic grade.

4.2. Chromatography

Precoated Merck silica gel F254 plates were used forTLC and spots were examined under UV light at254 nm and further visualized by sulfuric acid-anisalde-hyde spray. Column chromatography was performed onUetikon silica (0.2–0.06 mm).

4.3. Instruments and analyses

NMR spectra were obtained with a Varian Mercury300 MHz spectrometer. Chemical shifts are given inppm (d) relative to the residual solvent signal, in thecase of DMSO-d6 2.54 ppm for 1H and in the case ofCDCl3 7.26 ppm for 1H. All signals assigned to amino,amide hydrogen and hydroxyl groups were exchange-able with D2O. Exact mass measurements were per-formed on a quadrupole/orthogonal-acceleration time-of-flight (Q/oaTOF) tandem mass spectrometer (qToF2, Micromass, Manchester, UK) equipped with a stan-dard electrospray ionization (ESI) interface. Sampleswere infused in a 2-propanol/water (1:1) mixture at3 lL/min.

4.4. 9-(3-Amino-3-deoxy-b-DD-ribofuranosyl)-2-iodo-N6-methyladenine (1)

This compound was synthesized from 30 mg(0.069 mmol) of 25 by the procedure described forthe synthesis of 7; yield: 26 mg (92%). 1H NMR(300 MHz, DMSO-d6): d 1.66 (br s, 2H, NH2), 2.87(d, 3H, J = 4 Hz, N6-CH3), 3.41 (app t, 1H,J = 6.0 Hz, H3 0), 3.54–3.71 (m, 3H, H4 0 and H5 0Aand H5 0B), 4.10 (br s, 1H, 2 0-OH), 4.19 (dd, 1H,J = 2.6 and 4.4 Hz, H2 0), 4.98 (br s, 1H, 5 0-OH), 5.82(d, 1H, J = 2.6 Hz, H1 0), 8.11 (d, 1H, J = 4 Hz, N6H),8.30 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O):Calcd for C11H16N6O3I[M+H]+: 407.0330. Found:407.0332.

4.5. 9-(3-Amino-3-deoxy-b-DD-ribofuranosyl)-2-amino-N6-methyladenine (2)

This compound was synthesized from 27 (35 mg,0.11 mmol) by the procedure described for the syn-thesis of 7; yield: 30 mg (93%). 1H NMR(300 MHz, DMSO-d6): d 1.70 (br s, 2H, 3 0-NH2),2.86 (s, 3H, N6-CH3), 3.34 (app t, 1H, J = 6.0 Hz,H3 0), 3.52–3.70 (m, 3H, H4 0 and H5 0A andH5 0B), 4.16 (br s, 2H, 2 0-OH and H2 0), 5.16 (brs, 1H, 5 0-OH), 5.74 (d, 1H, J = 2.8 Hz, H1 0), 5.83(s, 2H, 2-NH2), 7.24 (br s, 1H, N6H), 7.91 (s,1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcdfor C12H18N7O3 [M+H]+: 296.1471. Found:296.1470.

1408 L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412

4.6. General procedure for the synthesis of alkynes 3, 4 and 5from 1

Compound 1 (50 mg, 0.12 mmol) was dissolved in Et3N(1.5 mL) and DMF (1 mL). After purging the solutionwith N2, (PPh3)2PdCl2 (8.6 mg, 0.012 mmol) and CuI(2.3 mg, 0.012 mmol) were added. The appropriate al-kyne (2 equiv) was subsequently added dropwise andthe mixture was stirred at room temperature for 3 h.The solvents were removed under reduced pressure,the residue was taken up in ethyl acetate and the solu-tion was filtered over a Celite pad. The residue remain-ing after solvent evaporation was purified on silica gelcolumn (CH2Cl2/MeOH, 90:10).

4.7. 9-(3-Amino-3-deoxy-b-DD-ribofuranosyl)-N6-methyl-2-phenylethynyladenine (3)

The reaction of 1 (50 mg, 0.12 mmol) with phenylacety-lene (27 lL, 0.24 mmol) gave compound 3 in 81% yield.1H NMR (300 MHz, DMSO-d6) d 2.07 (br s, 2H, 3 0-NH2), 2.94 (s, 3H, N6-CH3), 3.44 (app t, 1H, J = 5.3 Hz,H3 0), 3.56–3.75 (m, 3H, H4 0 and H5 0A and H5 0B), 4.09(br s, 1H, 2 0-OH), 4.24 (dd, 1H, J = 2.6 and 4.4 Hz,H2 0), 5.11 (br s, 1H, 5 0-OH), 5.94 (d, 1H, J = 2.3 Hz,H1 0), 7.46 (m, 3H, Ph), 7.61 (m, 2H, Ph), 7.95 (br s, 1H,N6H), 8.47 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd for C19H21N6O3[M+H]+: 381.1675. Found:381.1675.

4.8. 9-(3-Amino-3-deoxy-b-DD-ribofuranosyl)-N6-methyl-2-(4-methyl-phenyl)ethynyladenine (4)

The reaction of 1 (50 mg, 0.12 mmol) with 4-methyl-phenylacetylene (31 lL, 0.24 mmol) gave compound 4in 82% yield. 1H NMR (300 MHz, DMSO-d6) d 2.33(s, 3H, CH3Ph), 2.94 (s, 3H, N6-CH3), 3.50–3.79 (m,4H, H3 0, H4 0, H5 0A and H5 0B), 4.31 (dd, 1H, J = 2.4and 4.5 Hz, H2 0), 5.15 (br s, 1H, 5 0-OH), 5.74 (s, 1H,2 0-OH), 5.96 (d, 1H, J = 2.4 Hz, H1 0), 7.25 (m, 2H,Ph), 7.50 (m, 2H, Ph), 7.95 (br s, 1H, N6H), 8.46 (s,1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcdfor C20H23N6O3[M+H]+: 395.1831. Found: 395.1823.

4.9. 9-(3-Amino-3-deoxy-b-DD-ribofuranosyl)-N6-methyl-2-(1-hexyn-1-yl)adenine (5)

The reaction of 1 (50mg, 0.12 mmol)with 1-hexyn (28 lL,0.24 mmol) gave compound 5 in 80% yield. 1H NMR(300 MHz, DMSO-d6) d 0.90 (t, 3H, J = 7.03 Hz,CH2CH3), 1.37–1.54 (m, 4H, CH2CH2CH3), 2.41 (t, 2H,J = 7.04 Hz, C„CCH2), 2.88 (s, 3H, N6-CH3), 3.48–3.73 (m, 4H, H3 0, H4 0, H5 0A and H5 0B), 4.10–4.23 (m,2H, H2 0 and 2 0-OH), 5.10 (br s, 1H, 5 0-OH), 5.90 (d,1H, J = 2.5 Hz, H1 0), 7.83 (br s, 1H, N6H), 8.41 (s, 1H,H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd forC17H25N6O3[M+H]+: 361.1988. Found: 361.1982.

4.10. 9-(3-Amino-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-N6-methyl-2-iodoadenine (7)

Compound 13 (50 mg, 0.11 mmol) and PPh3 (57 mg,0.21 mmol) were dissolved in THF (2 mL). After stirring

for 10 min, H2O was added (270 lL, 15 lmol) and thereaction mixture was stirred for 2 days. The residue ob-tained after solvent evaporation was purified by columnchromatography (CH2Cl2/MeOH, 95:5) to yield 82% ofcompound 7. 1H NMR (300 MHz, DMSO-d6): d 1.79 (s,2H, 3 0-NH2), 2.69 (d, 3H, J = 4.7 Hz, CH3NHCO), 2.88(d, 3H, J = 4.1 Hz, N6-CH3), 3.54 (t, 1H, J = 11.1 Hz,H3 0), 4.10 (d, 1H, J = 6.2 Hz, H4 0), 4.30 (app t, 1H,J = 8.8 Hz, H2 0), 5.91 (br s, 1H, 2 0-OH), 5.93 (d, 1H,J = 3.81 Hz, H1 0), 8.03 (d, 1H, J = 4,7 Hz, NHCO),8.13 (d, 1H, J = 4.1 Hz, N6H), 8.48 (s, 1H, H8); ExactMass (ESI-MS, i-PrOH/H2O): Calcd for C12H17ClI-N7O3[M+H]+: 434.0439. Found: 434.0445.

4.11. 9-(3-Amino-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-2-amino-N6-methylpurine (8)

Compound 28 (40 mg, 0.12 mmol) and PPh3 (66 mg,0.25 mmol) were dissolved in THF (2 mL). After stirringfor 10 min, H2O was added (310 lL, 17 mmol) and themixture was stirred for 2 days. The residue obtained aftersolvent evaporation was purified by column chromatog-raphy (CH2Cl2/MeOH, 80:20) to give compound 8 in80% yield. 1H NMR (300 MHz, DMSO-d6): d 2.95 (br s,2H, 3 0-NH2), 2.66 (d, 3H, J = 4.7 Hz, CH3NHCO), 2.86(s, 3H, N6-CH3), 3.54 (app t, 1H, J = 5.4 Hz, H3 0), 4.05(d, 1H, J = 5.6 Hz, H4 0), 4.37 (app t, 1H, J = 4.7 Hz,H2 0), 5.82 (d, 1H, J = 4.1 Hz, H1 0), 5.88 (s, 2H, 2-NH2),7.31 (s, 1H, N6H), 8.04 (s, 1H, H8), 8.27 (d, 1H,J = 4.4 Hz, NHCO); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd for C12H19N8O3[M+H]+: 323.1580. Found:323.1579.

4.12. 9-(3-Amino-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-N6-methyl-2-phenylethynyladenine (9)

This compound was synthesized by the procedure de-scribed for the synthesis of 7 from 20 mg (0.046 mmol)of 14 in 96% yield (18 mg). 1H NMR (300 MHz,DMSO-d6): d 1.78 (s, 2H, 3 0-NH2), 2.71 (d, 3H,J = 4.4 Hz, CH3NHCO), 2.94 (s, 3H, N6-CH3), 4.33 (dd,1H, J = 5.0 and 5.3 Hz, H3 0), 4.13 (d, 1H, J = 5.3 Hz,H4 0), 4.36 (br s, 1H, H2 0), 5.94 (s, 1H, 2 0-OH), 6.02 (d,1H, J = 4.40, H1 0), 7.46 (m, 3H, Ph), 7.61 (m, 2H, Ph),8.04 (d, 1H, J = 4.4 Hz, N6H), 8.41 (d, 1H, J = 4.7 Hz,NHCO), 8.62 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd for C20H22N7O3[M+H]+: 408.1784.Found: 408.1787.

4.13. 9-(2-O-Acetyl-3-azido-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-2-amino-6-chloropurine (11)

4.13.1. Silylation of the base. 2-Amino-6-chloropurine(462 mg, 2.7 mmol) was treated with 1,1,1,3,3,3-hexam-ethyldisilazane (HMDS, 40 mL) and (NH4)2SO4

(0.27 mmol, 36 mg) and refluxed for 20 h. The silylatedcompound was concentrated and used without furtherpurification.

4.13.2. Vorbruggen coupling. 3-Azido-3-deoxy-1,2-di-O-acetyl-a-DD-ribofuronamide (10, 600 mg, 2.09 mmol) indry 1,2-dichloroethane (25 mL) was added to the silylated

L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412 1409

2-amino-6-chloropurine (462 mg, 2.7 mmol). The solu-tion was gently refluxed, and after 5 min TMSOTf(417lL, 2.3 mmol) was added dropwise. After 4 h, themixture was cooled to room temperature, quenched witha cold saturated NaHCO3 solution (80 mL) and extract-ed with CH2Cl2 (40 mL). The organic layer was driedwith MgSO4, filtered and evaporated to dryness. Theresidue was purified by column chromatography(CH2Cl2/MeOH, 98:2) to give 650 mg (79%) of com-pound 11. 1H NMR (300 MHz, CDCl3): d 2.15 (s, 3H,CH3CO), 2.90 (d, 3H, J = 5.0 Hz, CH3N), 4.53 (d, 1H,J = 3.2 Hz, H4 0), 4.90 (dd, 1H, J = 3.5 and 5.0 Hz,H3 0), 5.14 (s, 2H, NH2), 5.94 (t, 1H, J = 5.3 Hz, H2 0),5.97 (d, 1H, J = 6.2 Hz, H1 0), 7.09 (br s, 1H, NHCO),7.82 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O):Calcd for C13H15ClN9O4[M+H]+: 396.0935. Found:396.0932.

4.14. 9-(2-O-Acetyl-3-azido-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-6-chloro-2-iodopurine (12)

Isoamylnitrite (681 lL, 4.98 mmol) was added to amixture of 11 (650 mg, 1.65 mmol), I2 (418 mg,1.65 mmol), CH2I2 (1.37 mL, 16.5 mmol) and CuI(330 mg, 1.72 mmol) in 15 mL THF. The mixturewas refluxed for 45 min and then cooled to room tem-perature. Insoluble materials were removed by filtra-tion, and the filtrate was concentrated to dryness.The residue was purified by means of a silica gel col-umn, which was washed with CH2Cl2 until the iodinecolour disappeared and then eluted with CH2Cl2/MeOH, 98:2. Compound 12 was obtained in 79%yield. 1H NMR (300 MHz, CDCl3): d 2.13 (s, 3H,CH3CO), 3.06 (d, 3H, J = 4.98 Hz, CH3N), 4.56 (d,1H, J = 2.9 Hz, H4 0), 4.80 (dd, 1H, J = 2.9 and5.6 Hz, H3 0), 5.75 (dd, 1H, J = 5.9 and 7.0 Hz, H2 0),6.06 (d, 1H, J = 7.3 Hz, H1 0), 7.09 (br s, 1H, NHCO),8.11 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O):Calcd for C13H13ClIN8O4[M+H]+: 506.9794. Found:506.9822.

4.15. 9-(3-Azido-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-N6-methyl-2-iodoadenine (13)

Compound 12 (460 mg, 0.91 mmol) was dissolved inEtOH (10 mL). Methylammonium chloride (92 mg,1.36 mmol) and Et3N (158 lL, 1.13 mmol) were added,and the solution was refluxed overnight. The mixturewas concentrated to dryness, dissolved in 7 N NH3 inmethanol and stirred at room temperature for 2 h todeprotect the 2 0-hydroxyl group. The volatiles were re-moved under reduced pressure, and the residue waspurified by silica gel column (CH2Cl2/MeOH, 98:2).The product, 13, was realized, in 77% yield. 1H NMR(300 MHz, DMSO-d6): d 2.71 (d, 3H, J = 4.7 Hz,CH3NHCO), 2.89 (d, 3H, J = 4.4 Hz, N6-CH3), 4.33(d, 1H, J = 3.5 Hz, H4 0), 4.47 (dd, 1H, J = 3.5 and5.0 Hz, H3 0), 4.92 (dd, 1H, J = 5.6 and 11.4 Hz, H2 0),5.89 (d, 1H, J = 6.5 Hz, H1 0), 6.33 (d, 1H, J = 5.6 Hz,2 0-OH), 8.10 (d, 1H, J = 4,7 Hz, NHCO), 8.17 (d, 1H,J = 4.7 Hz, N6H), 8.36 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd for C12H15IN9O3[M+H]+:460.0344. Found: 460.0350.

4.16. 9-(3-Azido-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-N6-methyl-2-phenylethynyladenine (14)

A solution of compound 20 (30 mg, 0.06 mmol) and10 mL of 7 NNH3 in methanol was kept at room temper-ature for 2 h to allow deprotection of the 2 0-hydroxylgroup. Themixturewas concentrated to dryness andpuri-fied on a silica gel column (CH2Cl2/MeOH, 97:3). A 91%yield of compound 14was obtained. 1HNMR (300 MHz,DMSO-d6): d 2.73 (d, 3H, J = 3.5 Hz, CH3NHCO), 2.96(s, 3H, N6-CH3), 4.33 (d, 1H, J = 2.6 Hz, H4 0), 4.51 (dd,1H, J = 3.0 and 5.27 Hz, H3 0), 4.96 (app t, 1H,J = 5.3 Hz, H2 0), 5.98 (d, 1H, J = 6.7 Hz, H1 0), 7.46 (m,3H, Ph), 7.64 (m, 2H, Ph), 8.12 (s, 1H, N6H), 8.51 (s,1H, H8), 8.53 (d, 1H, J = 4.4 Hz, NHCO); Exact Mass(ESI-MS, i-PrOH/H2O): Calcd for C20H20N9O3[M+H]+:434.1688. Found: 434.1686.

4.17. Attempted synthesis of compound 14

Attempted conversion of 13 to 14 using the procedure asdescribed for 3, 4 and 5 failed to give 14, but resulted inthe formation of triazole 15 as the sole reaction product.1H NMR (300 MHz, DMSO-d6): d 2.78 (d, 3H,J = 4.6 Hz, N6-CH3), 3.0 (br s, 3H, CH3NHCO), 5.15(app q, 1H, J = 6.3 Hz, H2 0), 5.17 (d, 1H, J = 3.4 Hz,H4 0), 5.61 (dd, 1H, J = 3.6 and 6.3 Hz, H3 0), 6.26 (d,1H, J = 5.4 Hz, 2 0OH), 6.27 (d, 1H, J = 6.3 Hz, H1 0),7.36 (t, 1H, J = 7.3 Hz, 400-Ph), 7.46–7.51 (m, 2H, 400-Ph and 3H, C„CPh), 7.65 (br d, 2H, J = 6.8 Hz,C„CPh), 7.90 (d, 2H, J = 7.6 Hz, 400-Ph), 8.59 (s, 1H,H8), 8.71 (m, 2H, H500 and NHCO); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd for C28H25N9O3[M+H]+:536.2158. Found: 536.2166.

4.18. 2-Iodo-N6-methyl-(9-tetrahydropyran-2-yl)adenine(17)

Methylammonium chloride (28 mg, 0.41 mmol) andDMAP (67 mg, 0.54 mmol) were added to 100 mg(0.27 mmol) of compound 16 in EtOH (6 mL), and thesolution was refluxed overnight. The mixture was con-centrated to dryness and the residue was purified on asilica gel column (pentane/ethyl acetate, 50:50). The titlecompound was obtained in 81% yield. 1H NMR(300 MHz, DMSO-d6): d 1.52 (m, 6H, H2A 0 andH2B 0, H3A 0 and H3B 0, H4A 0 and H4B 0), 2.87 (s, 3H,NH CH3), 3.68 (t, 1H, J = 11.4 Hz, H5 0A), 3.96 (appd, 1H, J = 11.4 Hz, H5 0B), 5.43 (dd, 1H, J = 2.3 and10.9 Hz, H1 0), 7.26 (br s, 1H, NHCH3), 7.87 (s, 1H,H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd forC11H15N5O3I[M+H]+: 360.0323. Found: 360.0333.

4.19.N6-Methyl-2-phenylethynyl-tetrahydropyranyladenine(18)

Compound 17 (200 mg, 0.557 mmol) was dissolved in amixture of Et3N (3 mL) and DMF (1 mL) and the solu-tion was purged with N2. (PPh3)2PdCl2 (39 mg,0.056 mmol) and CuI (10.6 mg, 0.056 mmol) were added.Phenyl acetylene (112 lL, 1.11 mmol) was subsequentlyadded dropwise, and the mixture was stirred at roomtemperature for 3 h. The solvents were removed under

1410 L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412

reduced pressure, the residuewas taken up in ethyl acetateand the resulting solutionwas filtered over a pad ofCelite.After solvent evaporation, the residue was purified on asilica gel column (pentane/ethyl acetate, 50:50) to givecompound 18 in 91% yield. 1HNMR (300 MHz, CDCl3):d 1.18–2.25 (m, 6H, H2A 0 and H2B 0, H3A 0 and H3B 0,H4A 0 and H4B 0), 3.21 (s, 3H, NH CH3), 3.72 (app t,1H, J = 11.4 Hz, H5 0A), 3.96 (app d, 1H, J = 10.3 Hz,H5 0B), 5.79 (d, 1H, J = 9.1 Hz, H1 0), 7.20 (s, 1H, H8),7.30 (d, 3H, J = 5.3 Hz, H-Ph), 7.61 (m, 2H,H-Ph); ExactMass (ESI-MS, i-PrOH/H2O): Calcd for C19H20N5

O[M+H]+: 334.1667. Found: 334.1671.

4.20. N6-Methyl-2-phenylethynylpurine (19)

To a solution of 18 (170 mg, 0.510 mmol) in 10 mL ofCH2Cl2 was added slowly a solution of 0.78 mL TFA(10.2 mmol) and 0.78 mL CH2Cl2. After stirring at roomtemperature for 1 h, the solvent was evaporated, and theresidue was taken up in ethyl acetate, and the solutionwas washed with 7% NaHCO3. After silica gel chroma-tography (CH2Cl2/MeOH, 97:3), pure 19 was obtainedin a 72% yield. 1H NMR (300 MHz, DMSO): d 2.97(s, 3H, NH CH3), 7.30 (m, 3H, H-Ph), 7.61 (m, 2H,H-Ph), 7.88 (br s, 1H, NHCH3), 8.24 (s, 1H, H8); ExactMass (ESI-MS, i-PrOH/H2O): Calcd forC14H12N5[M+H]+: 250.1092. Found: 250.1073.

4.21. 9-(2-Acetyl-3-azido-3-deoxy-5-methylcarbamoyl-b-DD-ribofuranosyl)-N6-methyl-2-phenylethynyladenine (20)

Toamixture of 19 (150 mg, 0.602 mmol) andmethyl 3-az-ido-3-deoxy-1,2-di-O-acetyl-a-DD-ribofuronamide (10)(207 mg, 0.722 mmol) in 3 mL CH3CN were successivelyadded 223 lL (0.903 mmol) of N,O-bis(trimethylsilyl)-acetamide (BSA) and 131 lL (0.722 mmol) TMSOTf.The suspension was refluxed for 10 h. After being cooledto room temperature, the reaction was quenched with 7%NaHCO3 and extracted with CH2Cl2. The organic layerwaswashedwithbrine, filtered through a short padofCel-ite and evaporated to dryness. The crude material waspurified by column chromatography (CH2Cl2/MeOH,99:1), and compound 20 was obtained in 24.4% yield.1H NMR (300 MHz, DMSO-d6): d 2.12 (s, 3H, CH3CO),3.02 (d, 3H, J = 4.69 Hz, CH3NHCO), 3.25 (s, 3H, N6-CH3), 4.58 (d, 1H, J = 2.6 Hz, H4 0), 4.78 (dd, 1H,J = 2.1 and 5.3 Hz, H3 0), 5.78 (dd, 1H, J = 7.3 and12.9 Hz, H2 0), 5.92 (s, 1H, N6H) 6.02 (d, 1H, J = 7.6 Hz,H1 0), 7.40 (app d, 3H, J = 7.0 Hz H-Ph), 7.63 (app d,2H, J = 9.1 Hz, H-Ph), 7.81 (s, 1H, H8), 8.83 (s, 1H,NHCO); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd forC22H22N9O4[M+H]+: 476.1794. Found: 476.1800.

4.22. 9-(2-Acetyl-3-azido-3-deoxy-5-O-toluoyl-b-DD-ribofuranosyl)-2-amino-6-chloropurine (22)

To a mixture of 2-amino-6-chloropurine (90 mg, 0.53mmol) and 3-azido-3-deoxy-1,2-di-O-acetyl-5-O-toluo-yl-a-DD-ribofuranose (21) (240 mg, 0.64 mmol) in 3 mLCH3CN were successively added 196 lL (0.79 mmol)BSA and 115 lL (0.64 mmol) TMSOTf. The suspensionwas heated at 80 �C for 3 h. After being cooled to roomtemperature, the reaction was quenched with 7% NaH-

CO3 and extracted with CH2Cl2. The organic layer waswashed with brine, filtered through a short Celite padand evaporated to dryness. The residue was purified bycolumn chromatography (CH2Cl2/MeOH, 99:1) to yield200 mg (65%) of compound 22. 1H NMR (300 MHz,CDCl3): d 2.19 (s, 3H, CH3CO), 2.41(s, 3H, CH3-Ph),4.40 (m, 1H, H4 0), 4.52–4.58 (m, 1H, H5 0A 0), 4.76–4.82(m, 2H, H3 0 and H5 0B), 5.13 (br s, 2H, NH2), 5.94(d, 1H, J = 3.8 Hz, H1 0), 5.80 (dd, 1H, J = 3.8 and5.6 Hz, H2 0), 7.23 (d, 2H, J = 7.63 Hz, Ph), 7.79 (s,1H, H8), 7.86 (d, 2H, J = 8.2 Hz, Ph); Exact Mass (ESI-MS, i-PrOH/H2O): Calcd for C20H19N8O5Cl[M+H]+:487.1245. Found: 487.1246.

4.23. 9-(2-Acetyl-3-azido-3-deoxy-5-O-toluoyl-b-DD-ribofuranosyl)-6-chloro-2-iodopurine (23)

This compound was prepared by the procedure describedfor the synthesis of 12 from 22 (200 mg, 0.41 mmol); yield:200 mg (81%). 1HNMR (300 MHz, DMSO-d6): d 2.15 (s,3H, CH3CO), 2.35 (s, 3H, CH3-Ph), 4.37 (dd, 1H, J = 4.1and 7.6 Hz, H4 0), 4.52 (dd, 1H, J = 4.4 and 12.3 Hz,H5 0A 0), 4.65 (dd, 1H, J = 3.2 and 12.3 Hz, H5 0B), 4.96(dd, 1H, J = 5.8 and 7.6 Hz, H3 0), 6.03 (dd, 1H, J = 2.6and 5.28 Hz, H2 0), 6.29 (d, 1H, J = 2.9 Hz, H1 0), 7.24(d, 2H, J = 7.9 Hz, Ph), 7.68 (d, 2H, J = 8.2 Hz, Ph),8.74 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O):Calcd forC20H17N7O5ClINa[M+Na]+: 619.9924. Found:619.9920.

4.24. 9-(3-Azido-3-deoxy-2-hydroxyl-5-O-toluoyl-b-DD-ribofuranosyl)-2-iodo-N6-methyladenine (24)

The title compound was prepared as described for thesynthesis of 13 from 23 (200 mg, 0.335 mmol); yield:127 mg (69%). 1H NMR (300 MHz, DMSO-d6): d 2.36(s, 3H, CH3-Ph), 2.87 (d, 3H, J = 3.5 Hz, N6-CH3),4.29 (dd, J = 5.3 and 9.7 Hz, H4 0), 4.46–4.60 (m, 3H,H3 0, H5 0A and H5 0B), 5.01 (t, 1H, J = 4.7 Hz, H2 0),5.87 (d, 1H, J = 4.7 Hz, H1 0), 6.43 (d, 1H, J = 5.3 Hz,2 0-OH), 7.29 (d, 2H, J = 8.2 Hz, Ph), 7.78 (d, 2H,J = 8.2 Hz, Ph), 8.17 (d, 1H, J = 4.4 Hz, N6H), 8.21 (s,1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcdfor C19H20N8O4I [M+H]+: 551.0653. Found: 551.0649.

4.25. 9-(3-Azido-3-deoxy-b-DD-ribofuranosyl)-2-iodo-N6-methyl-adenine (25)

Ester 24 (127 mg, 0.23 mmol) was dissolved in 2.5 mLMeOH. Na� (11.28 mg, 0.32 mmol) was added and themixture was stirred at room temperature for 1 h. Thereaction was quenched by adding a mixture ofCH3COOH/H2O (9:1, v/v) to pH 7. The solution wasconcentrated to dryness, and the residue was purifiedby column chromatography (CH2Cl2/MeOH, 98:2) toyield 100 mg (95%) of compound 25. 1H NMR(300 MHz, DMSO-d6): d 2.88 (d, 3H, J = 4 Hz, N6-CH3), 3.52–3.67 (m, 2H, H5 0A and H5 0B), 3.94 (dd,1H, J = 7.26 and 3.81 Hz, H4 0), 4.28 (app t, 1H,J = 4.5 Hz, H3 0), 4.89 (app t, 1H, J = 5.4 Hz, H2 0),5.20 (br s, 1H, 5 0-OH), 5.80 (d, 1H, J = 6.16 Hz, H1 0),6.25 (br s, 1H, 2 0-OH), 8.18 (d, 1H, J = 4 Hz, N6H),8.28 (s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O):

L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412 1411

Calcd for C11H14N8O3I [M+H]+: 433.0235. Found:433.0237.

4.26. 9-(3-Azido-3-deoxy-5-O-toluoyl-b-DD-ribofuranosyl)-2-amino-N6-methyl-adenine (26)

Derivative 22 (100 mg, 0.21 mmol) was solubilized inEtOH (5 mL). Methylammonium chloride (35 mg,0.515 mmol) and Et3N (72 lL, 0.515 mmol) were added,and the solution was refluxed overnight. The mixturewas concentrated to dryness, the residue redissolved inmethanolic NH3 and the solution stirred at room temper-ature for 2 h to allow deprotection of the 2 0-hydroxylgroup. Themixturewas concentrated to dryness and puri-fied on a silica gel column (CH2Cl2/MeOH, 97:3). Com-pound 26 was obtained in 68% yield. 1H NMR(300 MHz, DMSO-d6): d 2.36 (s, 3H, CH3-Ph), 2.84 (brs, 3H, N6-CH3), 4.23 (dd, J = 5.27 and 9.38 Hz, H4 0),4.43 (dd, 1H, J = 5.6 and 11.6, H5 0A), 4.55 (m, 2H,H3 and H5 0B), 4.48 (app t, 1H, 4.7 Hz, H2 0), 5.78(d, 1H, J = 4.7 Hz, H1 0), 5.95 (br s, 2H, 2-NH2), 6.34(d, 1H, J = 7.9 Hz, 2 0-OH), 7.28 (br s, 1H, N6H), 7.29(d, 2H, J = 7.9 Hz, Ph), 7.79 (s, 1H, H8), 7.82 (d, 2H,J = 8.2 Hz, Ph); Exact Mass (ESI-MS, i-PrOH/H2O):Calcd for C19H22N9O4 [M+H]+: 440.1794. Found:440.1789.

4.27. 9-(3-Azido-3-deoxy-b-DD-ribofuranosyl)-2-amino-N6-methyladenine (27)

The title compound was synthesized from 26 (60 mg,0.013 mmol) by the procedure described for the synthe-sis of 25; yield: 40 mg (91%). 1H NMR (300 MHz,DMSO-d6): d 2.85 (br s, 3H, N6-CH3), 3.48–3.64 (m,2H, H5 0A and H5 0B), 3.89 (dd, 1H, J = 3.2 and7.0 Hz, H4 0), 4.25 (dd, 1H, J = 3.2 and 5.6 Hz, H3 0),4.88 (app t, 1H, J = 5.3 Hz, H2 0), 5.58 (t, J = 6.8 Hz,1H, 5 0-OH), 5.71 (d, 1H, J = 6.2 Hz, H1 0), 5.85 (s, 2H,NH2), 6.16 (s, 1H, 2 0-OH), 7.31 (br s, 1H, N6H), 7.89(s, 1H, H8); Exact Mass (ESI-MS, i-PrOH/H2O): Calcdfor C11H16N9O3[M+H]+: 322.1376. Found: 322.1365.

4.28. 9-[3-Azido-3-deoxy-5-(methylcarbamoyl)-b-DD-ribofuranosyl]-2-amino-6-chloropurine (28)

Derivative 11 (120 mg, 0.30 mmol)was dissolved inEtOH(6 mL). Methylammonium chloride (30 mg, 0.46 mmol)and Et3N (53 lL, 0.38 mmol) were added, and the solu-tion was refluxed overnight. The mixture was concentrat-ed to dryness, the remaining solid dissolved in 7 NNH3 inmethanol and the solution stirred at room temperaturefor 2 h to deprotect the 2 0-hydroxyl group. The mixturewas concentrated to dryness and the residue was purifiedon a silica gel column (CH2Cl2/MeOH, 95:5). Compound28 was obtained in 79% yield. 1H NMR (300 MHz,DMSO-d6): d 2.68 (d, 3H, J = 4.7 Hz, CH3NHCO), 2.87(s, 3H, N6-CH3), 4.24 (d, 1H, J = 2.9 Hz, H-4 0), 4.42(dd, 1H, J = 2.9 and 5.3 Hz, H3 0), 4.97 (dd, 1H, J = 5.6and 11.7 Hz, H2 0), 5.81 (d, 1H, J = 6.8 Hz, H1 0), 5.91 (s,2H, NH2), 6.26 (d, 1H, J = 5.3 Hz, 2 0-OH), 7.31 (s, 1H,N6H), 7.96 (s, 1H, H8), 8.35 (d, 1H, J = 4.7 Hz, NHCO);Exact Mass (ESI-MS, i-PrOH/H2O): Calcd forC12H17N10O3[M+H]+: 349.1484. Found: 349.1491.

4.29. Biological assays

4.29.1. Cell culture and membrane preparation. CHO cellsexpressing recombinant human A3ARs were cultured inDMEM (Dulbecco�s modified Eagle�s medium) and F12(1:1) supplemented with 10% foetal bovine serum,100 U/mL penicillin, 100 lg/mL streptomycin, 2 lmol/mL glutamine and 800 lL geneticin. After harvest andhomogenization, the cells were centrifuged at for10 min. The pellet was resuspended in 50 mM Tris–HClbuffer (pH 8.0) containing 10 mM MgCl2 and 1 mMEDTA. The suspensionwas homogenized with an electrichomogenizer for 10 s and was then recentrifuged at20,000g for 20 min at 4 �C. The resulting pellets wereresuspended in buffer containing 3 U/mL of adenosinedeaminase, and the suspensionwas stored at�80 �Cpriorto the binding experiments. The protein concentrationwas measured using the Bradford assay.25

4.29.2. Radioligand binding studies.For theA3ARbindingexperiments, the procedures were similar to those previ-ously described.26 Briefly, each tube contained 100 lLof membrane suspension, 50 lL [125I]I-AB-MECA (finalconcentration 0.5 nM) and 50 lL of increasing concen-trations of compounds in Tris–HCl buffer (50 mM, pH7.4) containing 10 mM MgCl2, 1 mM EDTA. Non-spe-cific binding was determined using 10 lM NECA. Themixtures were incubated at 25 �C for 60 min. Bindingreactions were terminated by filtration throughWhatmanGF/B filters under reduced pressure using a MT-24 cellharvester (Brandel, Gaithersburg, MD). Filters werewashed three times with ice-cold buffer. Radioactivitywas determined in aBeckman 5500B c-counter. The bind-ing of [3H]R-PIA to the recombinant hA1AR and thebinding of [3H]CGS21680 to the recombinant hA2AARwere performed as previously described.27,28

4.29.3. Cyclic AMP accumulation assay. Intracellular cyc-lic AMP levels were measured by the competitive proteinbindingmethod.29 CHO cells expressing recombinant hu-man30 ARs were harvested by trypsinization. After resus-pension in the medium, cells were plated in 24-well platesin 0.5 mL medium/well. After 24 h, the medium was re-moved and cells were washed three times with 1 mL/wellDMEM containing 50 mM N-2-hydroxyethylpiper-azine-N 0-2-ethanesulfonic acid, pH 7.4. Cells were thentreated with agonists and/or test compounds in the pres-ence of rolipram (10 lM) and adenosine deaminase(3 U/mL) and incubated at 37 �C. For A3AR, after 45min forskolin (10 lM) was added to the medium, andincubation was continued for an additional 15 min. Thereaction was terminated upon removal of the medium,and the cells were lysed with 200 lL/well of 0.1 M ice-coldHCl. The cell lysate was resuspended and stored at�20 �C. For determination of cyclic AMP production,protein kinase A (PKA) was incubated with [3H]cyclicAMP (2 nM) in K2HPO4/EDTA buffer (K2HPO4,150 mM; EDTA, 10 mM), 20 lL of the cell lysate and30 lL of 0.1 M HCl. Bound radioactivity was separatedby rapid filtration through Whatman GF/C filters underreduced pressure and washed once with cold buffer.Bound radioactivity was subsequently measured by scin-tillation spectrometry.

1412 L. Cosyn et al. / Bioorg. Med. Chem. 14 (2006) 1403–1412

References and notes

1. Muller, C. Curr. Top. Med. Chem. 2000, 7, 1269.2. Poulsen, S. A.; Quinn, R. J. Bioorg. Med. Chem. 1998, 6,

619.3. Liang, B. T.; Jacobson, K. A. Proc. Natl. Acad. Sci.

U.S.A. 1998, 95, 6995.4. von Lubitz, D. K.; Lin, R. C.; Popik, P.; Carter,

M. F.; Jacobson, K. A. Eur. J. Pharmacol. 1994,263, 59.

5. Jacobson, K. A.; Moro, S.; Kim, Y. C.; Li, A. H. DrugDev. Res. 1998, 45, 113.

6. Brambilla, R.; Cattabeni, F.; Ceruti, S.; Barbieri, D.;Franceshi, C.; Kim, Y.; Jacobson, K. A.; Klotz, K. N.;Lohse, M. J.; Abbracchio, M. P. Naunyn-Schmiedeberg�sArch. Pharmacol. 2000, 361, 225.

7. Beaven, M. A.; Ramkumar, V.; Ali, H. Trends Pharmacol.Sci. 1994, 15, 13.

8. von Lubitz, D. K.; Carter, M. F.; Deutsch, S. I.; Lin, R.C.; Mastropaolo, J.; Meshulam, Y.; Jacobson, K. A. Eur.J. Pharmacol. 1995, 275, 23.

9. Ramkumar, V.; Stiles, G.; Beaven, M. A.; Ali, H. J. Biol.Chem. 1993, 268, 16887.

10. Okamura, T.; Kurogi, Y.; Hashimoto, K.; Sato, S.;Nishikawa, H.; Kiryu, K.; Nagao, Y. Bioorg. Med. Chem.Lett. 2004, 3775.

11. Civan, M. M.; Macknight, A. D. C. Exp. Eye Res. 2004,78, 625.

12. Van Tilburg, E.; von Frijtag Drabbe Kunzel, J.; deGroote, M. J. Med. Chem. 2002, 45, 420.

13. Jacobson, K. A.; Knutsen, L. In Handbook of Experimen-tal Pharmacology; Abbracchio, M., Williams, M., Eds.;Springer, A.: Berlin, 2001; Vol. 151/I, p 302.

14. Volpini, R.; Costanzi, S.; Lambertucci, C.; Vittori, S.;Lorenzen, A.; Klotz, K.-N.; Cristalli, G. Bioorg. Med.Chem. Lett. 2001, 11, 1931.

15. Costanzi, S.; Lambertucci, C.; Vittori, S.; Volpini, R.;Cristalli, G. J. Mol. Graph. Model 2003, 21, 253.

16. Volpini, R.; Constanzi, S.; Lambertucci, C.; Vittori, S.;Cristalli,G.Nucleosides,NucleotidesNucleicAcids 2001, 20,775.

17. DeNinno, M. P.; Masamune, H.; Chenard, L. K.; DiRico,K. J.; Eller, C.; Etienne, J. B.; Tickner, J. E.; Kennedy, S.P.; Knight, D. R.; Kong, J.; Oleynek, J. J.; Tracey, W. R.;Hill, R. J. J. Med. Chem. 2003, 46, 353.

18. Jeong, L. S.; Kim, M. J.; Gao, Z.-G.; Kim, S.-K.;Jacobson, K. A.; Chun, M. W. Bioorg. Med. Chem. Lett.2004, 14, 4851.

19. Volpini, R.; Constanzi, S.; Lambertucci, C.; Taffi, S.;Vittori, S.; Klotz, K.-N.; Cristalli, G. J. Med. Chem. 2002,45, 3271.

20. Mathieu, R.; Baurand, A.; Schmitt, M.; Gachet, C.;Bourguignon, J.-J. Bioorg. Med. Chem. 2004, 12, 1769.

21. Huisgen, R. In 1,3—Dipolar Cycloaddition Chemistry;Padwa, A., Ed.; Wiley: New York, 1984; pp 1–176.

22. Brun, V.; Legraverend, M.; Grierson, D. S. TetrahedronLett. 2001, 42, 8161.

23. Vorbruggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber.1981, 114, 1234.

24. Caddel, J.; Chapman, A.; Cooley, B.; Downey, B.;LeBlanc, M.; Jackson, M.; O�Connell, T.; Roper, T.;Xie, S. J. Org. Chem. 2004, 69, 3212.

25. Bradford, M. M. Anal. Biochem. 1976, 72, 248.26. Gao, Z.-G.; Kim, S.-K.; Biadatti, T.; Chen, W.; Lee, K.;

Barak, D.; Kim, S. G.; Johnson, C. R.; Jacobson, K. A. J.Med. Chem. 2002, 45, 4471.

27. Gao, Z.-G.; Blaustein, J. B.; Gross, A. S.; Melman, N.;Jacobson, K. A. Biochem. Pharmacol. 2003, 65, 1675.

28. Hutchison, A. J.; Williams, M.; de-Jesus, R.; Yokoyama,R.; Oei, H. H.; Ghai, G. R.; Webb, R. L.; Zoganas, H. C.;Stone, G. A.; Jarvis, M. F. J. Med. Chem. 1990, 33, 1919.

29. Nordstedt, C.; Fredholm, B. B. Anal. Biochem. 1990, 189,231.

30. Salvatore, C.A.; Jacobson,M.A.; Taylor,H. E.; Linden, J.;Johnson, R. G. Proc. Natl. Acad. Sci. U.S.A. 1993,90, 10365.